The robot is a camera-equipped catheter, similar to the kind that are already commonly used during conventional surgery. Normally, those are controlled manually by the surgeon in order to navigate to a specific part of the body, like a heart valve. In this case, however, the robot is able to autonomously guide itself to the desired location. First, it’s given an anatomical map and data from imaging before the operation. Then it processes the video feed from its camera with machine learning to perform wall-following inside the body until it reaches its destination.
To test how well it worked, the team performed a very difficult procedure called a paravalvular aortic leak closure on pigs, which repairs leaks around replacement heart valves. In their trials, the robot was able to reach the valve in roughly the same amount of time as a surgeon would. Once there, the surgeon was able to take over and perform the delicate work. The advantages of that are profound. It would reduce surgeons’ work load and fatigue, and also could make remote surgery more efficient. The process of navigating the body could already be completed, and a remote surgeon could simply step in to finish the operation — allowing them to do many more each day.